🌾 KrishiSathi AI - Bharat Insight Studio

Empowering Public Data Analysis with AI & Multi-Language Support

A modern, open-source AI-powered analytics platform designed for exploring public datasets with advanced visualization, multi-language insights, and actionable recommendations. Built on top of Google Gemini AI, it combines powerful analytics with accessibility.

Key Features

AI-Powered Intelligence

Google Gemini Integration - Advanced AI models (3.5 Flash & 3.1 Pro)
Smart Analysis - Automatic insights, trends, anomalies, and risk assessment
Customizable Prompts - Pre-built templates or create your own
Confidence Scoring - Track AI confidence on each analysis

Multilingual Capabilities

Insights available in 12+ Indian languages:

English, Hindi (हिन्दी), Bengali (বাংলা)
Tamil (தமிழ்), Telugu (తెలుగు), Kannada (ಕನ್ನಡ)
Marathi (मराठी), Gujarati (ગુજરાતી), Malayalam (മലയാളം)
Punjabi (ਪੰਜਾਬਿ), Odia (ଓଡ଼ିଆ), Urdu (اردو)

Data Management

Multi-Format Support - CSV, JSON, Excel (XLSX)
Automatic Parsing - Smart type detection and data cleaning
Statistics Generation - Headers, duplicates, missing values, distributions
Persistent Storage - Local JSON database

Visualization & Reporting

Interactive Charts - Recharts-powered visualizations
Report Generation - Download as Markdown files
Data Exploration - Trends, patterns, comparisons
Export Options - Multiple format support

Modern UI/UX

Dark Theme - Eye-friendly glass-morphism design
Responsive Layout - Optimized for desktop and tablet
Intuitive Navigation - Sidebar-based tab system
Real-time Updates - Hot module reloading

Quick Start

Prerequisites

Requirement	Version	Link
Node.js	v18+	Download
npm	v9+	Included with Node.js
Gemini API Key	-	Get API Key

Installation

Step 1: Clone & Navigate

cd KrishiSathi_AI

Step 2: Install Dependencies

npm install

Step 3: Configure API Key

Create .env.local file in the project root:

GEMINI_API_KEY=your_gemini_api_key_here

Step 4: Start Development Server

npm run dev

Step 5: Open in Browser

http://localhost:3000

Note: If port 3000 is busy, the server automatically uses the next available port (3001, 3002, etc.)

Available Scripts

Command	Description	Output
`npm run dev`	Development server with hot reload	Localhost dev mode
`npm run build`	Production build (Vite + esbuild)	`dist/` folder
`npm start`	Run production build	Compiled app
`npm run clean`	Remove build artifacts	Clean slate
`npm run lint`	TypeScript type checking	Type errors only

Project Structure

KrishiSathi_AI/
│
├── src/
│   ├── components/
│   │   ├── AIStudio.tsx              #  Main AI analysis interface
│   │   ├── DashboardHome.tsx         #  Home dashboard
│   │   ├── UploadPage.tsx            # File upload portal
│   │   ├── DatasetLibrary.tsx        # Dataset browser
│   │   ├── VisualizationStudio.tsx   # Chart visualization
│   │   ├── ReportGenerator.tsx       # Report builder
│   │   ├── HistoryTab.tsx            # ⏱Analysis history
│   │   ├── SettingsTab.tsx           # User settings
│   │   ├── ProfileTab.tsx            # User profile
│   │   ├── Navbar.tsx                # Top navigation
│   │   ├── Sidebar.tsx               # Main sidebar
│   │   ├── AILoading.tsx             # Loading state
│   │   ├── EmptyState.tsx            # Empty placeholders
│   │   └── PromptBuilder.tsx         # Custom prompt editor
│   ├── App.tsx                       # Main app component
│   ├── main.tsx                      # React entry point
│   ├── index.css                     # Global styles & theme
│   ├── types.ts                      # TypeScript interfaces
│   └── assets/
│       └── images/
│
├── data/
│   ├── db.json                       # Local data store
│   └── Daily Retail Price of Potato.csv
│
├── server.ts                         # Express backend
├── vite.config.ts                    # Vite configuration
├── tsconfig.json                     # TypeScript config
├── package.json                      # Dependencies
└── README.md                         # This file

🔌 API Reference

Base URL

http://localhost:3000/api

Datasets

Get All Datasets

GET /api/datasets

Response:

[
  {
    "id": "dataset_123",
    "name": "Potato Prices",
    "rowCount": 1000,
    "headers": ["Date", "Price", "Region"],
    "data": [...]
  }
]

Upload Dataset

POST /api/datasets
Content-Type: multipart/form-data

Parameters:

file (required) - CSV, JSON, or XLSX file
name (optional) - Custom dataset name

Delete Dataset

DELETE /api/datasets/:id

AI Analysis

Run Analysis

POST /api/ai/analyze
Content-Type: application/json

Request Body:

{
  "datasetId": "dataset_123",
  "promptText": "Find trends in this data",
  "promptName": "Find Trends",
  "language": "English",
  "temperature": 0.4,
  "model": "gemini-3.5-flash"
}

Response:

{
  "id": "analysis_456",
  "timestamp": "2024-01-15T10:30:00Z",
  "response": {
    "summary": "Executive summary...",
    "insights": ["Insight 1", "Insight 2"],
    "reasoning": "Analysis reasoning...",
    "evidence": "Supporting data...",
    "confidenceScore": 85,
    "limitations": "Data limitations...",
    "recommendedActions": ["Action 1", "Action 2"],
    "exploreQuestions": ["Question 1"]
  }
}

History

Get Analysis History

GET /api/history

Delete History Entry

DELETE /api/history/:id

Reports

Get All Reports

GET /api/reports

Create Report

POST /api/reports

Delete Report

DELETE /api/reports/:id

Settings & Profile

Get Settings

GET /api/settings

Update Settings

PUT /api/settings

Get Profile

GET /api/profile

Update Profile

PUT /api/profile

How to Use

1. Upload a Dataset

Click "Upload" in the sidebar
Select CSV, JSON, or Excel file
System auto-parses and validates data
Dataset appears in library

2. Analyze with AI

Go to "Gemma" (AI Studio)
Select dataset from dropdown
Choose prompt template or write custom
Adjust temperature (0.1 = factual, 1.0 = creative)
Select output language
Click "Initiate Audit Analysis"

3. Review Results

Executive summary
Key factual insights
Analytical reasoning
Evidence utilization
Risk warnings & gaps
Recommended actions

4. Export & Share

Copy - Copy full text to clipboard
Download - Get Markdown report

5. Track History

View all past analyses
Timestamps & models used
Re-run previous prompts

Supported File Formats

CSV

Date,Price,Region
2024-01-01,150,North
2024-01-02,152,South

✅ Automatic header detection
✅ Quote & escape handling
✅ Type inference (numeric, categorical, text)
✅ Null value detection

JSON

[
  {"Date": "2024-01-01", "Price": 150, "Region": "North"},
  {"Date": "2024-01-02", "Price": 152, "Region": "South"}
]

✅ Array of objects
✅ Nested structures
✅ Mixed types

Excel (XLSX)

Workbook → Sheet → Table

✅ Multiple sheets
✅ Formula values
✅ Cell formatting

AI Prompt Templates

10 Pre-built Templates

#	Template	Purpose	Creativity
1️⃣	Generate Summary	Dataset overview	0.3
2️⃣	Find Trends	Correlations & patterns	0.4
3️⃣	Find Anomalies	Outliers & spikes	0.3
4️⃣	Compare Categories	Class distributions	0.4
5️⃣	Risk Analysis	Policy risks & gaps	0.3
6️⃣	Recommendations	Strategic advice	0.5
7️⃣	Question Answering	Specific queries	0.4
8️⃣	Explain Dataset	Layman terms	0.4
9️⃣	Translate Results	Localized insights	0.2
🔟	Executive Report	Comprehensive paper	0.4

Create Custom Prompts

Settings → Custom Prompts
Name your template
Write system & user prompts
Set creativity level
Save for future use

Technology Stack

Frontend

React 19 - UI framework
TypeScript 5.8 - Type safety
Vite 6 - Lightning-fast build tool
Tailwind CSS 4 - Utility-first styling

Backend

Express.js - Server framework
Node.js - Runtime

AI & Data

Google Gemini API - AI models
XLSX - Excel parsing
Recharts - Data visualization

UI Components

Lucide React - Icon library
Motion - Animations

Development

tsx - TypeScript executor
esbuild - Bundler
ESLint - Linting

Data Storage

Local Database Structure

data/db.json
├── datasets[]         # Uploaded datasets
├── history[]          # Analysis history
├── reports[]          # Generated reports
├── customPrompts[]    # User templates
├── settings{}         # App configuration
└── profile{}          # User profile

Example Database Entry

{
  "datasets": [
    {
      "id": "dataset_abc123",
      "name": "Potato Prices 2024",
      "rowCount": 365,
      "uploadedAt": "2024-01-15T10:30:00Z",
      "headers": ["Date", "Price", "Region"],
      "statistics": {
        "columns": [...],
        "duplicateCount": 0,
        "totalMissing": 5
      },
      "data": [...]
    }
  ]
}

Troubleshooting

Port Already in Use

Error: listen EADDRINUSE: address already in use

Solution: Server auto-selects next port. Check console for actual URL.

Gemini API Key Error

Error: GEMINI_API_KEY is not defined

Solution:

Create .env.local file
Add: GEMINI_API_KEY=your_key_here
Restart server

CSS Import Error

@import must precede all other statements

Solution: Ensure @import statements are at top of src/index.css

Build Fails

# Clear and reinstall
npm run clean
npm install
npm run build

Hot Reload Issues

# Disable HMR temporarily
DISABLE_HMR=true npm run dev

🔒 Security & Privacy

✅ Local data storage (no cloud sync)
✅ No personal data collection
✅ API key stored locally only
✅ Datasets never shared
✅ Privacy mode available in settings

🤝 Contributing

We welcome contributions! Please:

Fork the repository
Create feature branch: git checkout -b feature/amazing-feature
Commit changes: git commit -m 'Add amazing feature'
Push branch: git push origin feature/amazing-feature
Open Pull Request

📝 License

This project is open-source and available under the MIT License. See LICENSE for details.

🙏 Acknowledgments

Google Gemini AI - Powering intelligent analysis
React & Vite - Modern frontend development
Tailwind CSS - Beautiful UI styling
Open Data Community - Supporting public data initiatives

Support & Contact

Email: support@krishisathi.com
Issues: GitHub Issues
Discussions: GitHub Discussions
Documentation: Full docs available in /docs

Roadmap

Built with ❤️ for public data analysis

⬆ Back to top

Name		Name	Last commit message	Last commit date
assets		assets
data		data
src		src
.gitignore		.gitignore
README.md		README.md
index.html		index.html
metadata.json		metadata.json
package-lock.json		package-lock.json
package.json		package.json
server.ts		server.ts
tsconfig.json		tsconfig.json
vite.config.ts		vite.config.ts

Latest commit

History

Folders and files

Repository files navigation

🌾 KrishiSathi AI - Bharat Insight Studio

Key Features

AI-Powered Intelligence

Multilingual Capabilities

Data Management

Visualization & Reporting

Modern UI/UX

Quick Start

Prerequisites

Installation

Step 1: Clone & Navigate

Step 2: Install Dependencies

Step 3: Configure API Key

Step 4: Start Development Server

Step 5: Open in Browser

Available Scripts

Project Structure

🔌 API Reference

Base URL

Datasets

Get All Datasets

Upload Dataset

Delete Dataset

AI Analysis

Run Analysis

History

Get Analysis History

Delete History Entry

Reports

Get All Reports

Create Report

Delete Report

Settings & Profile

Get Settings

Update Settings

Get Profile

Update Profile

How to Use

1. Upload a Dataset

2. Analyze with AI

3. Review Results

4. Export & Share

5. Track History

Supported File Formats

CSV

JSON

Excel (XLSX)

AI Prompt Templates

10 Pre-built Templates

Create Custom Prompts

Technology Stack

Frontend

Backend

AI & Data

UI Components

Development

Data Storage

Local Database Structure

Example Database Entry

Troubleshooting

Port Already in Use

Gemini API Key Error

CSS Import Error

Build Fails

Hot Reload Issues

🔒 Security & Privacy

🤝 Contributing

📝 License

🙏 Acknowledgments

Support & Contact

Roadmap

About

Resources

Stars

Watchers

Forks